Increased efficiency through pricing in online labor markets

We present a matching-pricing algorithm for online information labor markets motivated by exploratory study of Amazon Mechanical Turk (AMT). The algorithm addresses many of the challenges surfaced by the AMT study. We find that AMT pricing is largely done using rules of thumb, which results in inaccurate pricing. Such pricing leads to either excessive costs for employers or low income for workers and likely reduces the rate of tasks completed. We offer an improved mechanism based on a descending pricing function sensitive to the demand for work and the supply of workers. This approach grants employers the flexibility to design tasks to optimize cost and quality while empowering workers to select tasks that meet their occupation and income expectations. Additionally our pricing algorithm holds several unique advantages which include: motivating workers to quickly perform tasks, revealing workers' true market demand at corresponding prices, allowing workers to guarantee the assignment of complementary tasks assuring reduced cost performance, and bounding employers' total expense.